Schwendenmann,

Available on-line at:Mitchell: Carbon accumulation in an urban park

http://www.newzealandecology.org/nzje/ 213

Carbon accumulation by native trees and soils in an urban park, Auckland

Luitgard Schwendenmann* and Neil D. Mitchell

School of Environment, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand
*Author for correspondence (Email: l.schwendenmann@auckland.ac.nz)

Published online: 9 April 2014

Abstract: Carbon storage by trees and soil in urban areas is of increasing interest as a potential greenhouse
gas mitigation measure. Our objectives were to (1) quantify carbon accumulation in above- and below-ground
tree biomass, organic layer and mineral soil (0–5 cm) of a 27-year-old planted forest in Auckland and (2)
compare the sequestration potential of urban trees with natural shrublands and forests in New Zealand. A
mixed-species allometric equation for urban-grown native trees was developed on the basis of the tree biomass
of 21 trees belonging to four species (Corynocarpus laevigatus, Kunzea ericoides, Pittosporum eugenioides,
P. tenuifolium). Our allometric equation and a recently developed mixed-species equation for New Zealand
native forest species produced similar results. A total of 45.9 Mg C ha–1 was stored in above- and below-ground
tree biomass in our Auckland park, equating to an average annual carbon sequestration rate of 1.7 Mg C ha–1.
This rate is within the range reported for New Zealand shrublands. Despite the adverse conditions posed by
some urban environments (e.g. poor-quality soil, contamination), planted urban forests can sequester carbon
at similar rates to natural vegetation in New Zealand.
Keywords: allometric equations; carbon sequestration; planted forest; root biomass; tissue and wood chemistry

Introduction (0–5 cm) in this planted forest. To do this we developed an

Increased urbanisation is destroying natural ecosystems
and degrading the environment of urban areas. To mitigate
some of these problems local governments across the globe
conduct large tree-planting projects (i.e. urban afforestation)
to provide habitats and ecosystem services such as carbon (C)
sequestration, microclimate regulation, noise and air pollution
reduction, and stormwater attenuation (Auckland Council
2012; Gaffin et al. 2012; Oldfield et al. 2013).
Besides planted forests there are many small patches
of remnant forest scattered throughout urban areas in
New Zealand. For example, approximately 25% of Auckland’s
urban area consists of green spaces, dominated by urban
parkland and reserves (Auckland Council unpublished 2012
report ‘Urban Auckland green space GIS analysis’). All these
areas of woody and herbaceous vegetation take up CO2 for
photosynthesis. Urban green space cover varies widely among
cities but tends to be slightly higher in ‘sprawling’Australasian
and American cities compared with ‘compact’ European cities
(Fuller & Gaston 2009).
A higher proportion of vegetated surfaces, together with
favourable climatic conditions, may result in a higher potential
for C uptake in New Zealand cities. Greater growth of tree
seedlings in urban environments was reported by Searle et al.
(2012) suggesting urban trees can grow faster than trees in rural
settings. However, soils in urban areas are often compacted,
have low organic matter content, and may be contaminated (US
EPA 2011). Studies have shown that adverse soil conditions
and the exposure to pollutants and higher temperatures may
have a detrimental effect on tree growth and thus C uptake in
urban forests (e.g. Pouyat et al. 1995).
In this study we had the opportunity to investigate the
accumulation of carbon by trees and soils in a small, 27-year-
old, planted, urban forest in Auckland. Our first objective was
to quantify the amount of carbon accumulated in above- and
below-ground tree biomass, organic layer and mineral soil
New Zealand Journal of Ecology (2014) 38(2): 213-220 © New Zealand Ecological Society.
allometric equation for urban-grown native New Zealand
trees and compared this equation with allometric equations
commonly used to estimate tree carbon stocks in natural
forests in New Zealand. We also investigated the carbon
allocation pattern among species and our second objective was
to compare our results with the carbon storage/sequestration
values of native and exotic tree species growing in rural
New Zealand shrublands and forests. Our study contributes to
the understanding of carbon storage in New Zealand vegetation.
Materials and methods
Study site and vegetation survey
The study site was in a small urban park (Newmarket Park,
36°51'54"S; 174°47'2"E) located north-east of the Newmarket
suburb, Auckland City, New Zealand. The climate of Auckland
is characterised by warm humid summers and mild wet winters
with an average annual rainfall between 1000 and 1300 mm
(NIWA 2012). The section of park under study was a steep
(26°) south-east-facing slope at the edge of the park. This
area contained 1.87 ha of native trees that had been planted in
1983 as a ‘beautification’ project on the site of a landfill that
had been closed. The landfill was covered with a 3-m layer of
soil and sediments on in situ Waitemata sandstone. The infill
material was low in soil carbon (<0.3%) and nitrogen (<0.02%)
concentrations. Soils with soil organic matter contents <1% are
classified as ‘severely impacted’ and are considered unsuitable
for green infrastructure (i.e. trees and urban forestry; US EPA
2011) without soil amelioration measures such as mulching.
In 2010 the trees were cleared as the slope had to be
restabilised, presenting the opportunity to obtain some trees
for carbon analysis. We were only informed at short notice.
This resulted in a number of constraints imposed by our lack
of control over the research design. The vegetation survey
described below had been carried out by contractors prior to
214 New Zealand Journal of Ecology, Vol. 38, No. 2, 2014
clearance of the urban forest and at some time previous to
our involvement; by the time we reached the site much of it
had already been cleared. We were given a brief opportunity
to visit the site and identify trees we would like to analyse.
The trees were extracted and stockpiled to the contractors’
timetable and we were then given 3 days on which we could
work on them. No bulk storage facilities were available to us,
so all the basic measurement and sample collection was carried
out on site. This is also why it was not feasible to separate the
foliage from the branches.
The vegetation survey consisted of four transects parallel
to each other, from the top of the slope to the bottom. Transect
lengths varied from 35 to 50 m, depending upon the width
of the vegetation at that point. All trees within 1 m of either
side of the transect line were identified and their diameter at
1.4 m above ground measured. A general indication of tree
height was recorded.
Tree selection
As part of the project, the contractors agreed to dig up, in
their entirety, 21 trees of four native species to measure
the above- and below-ground biomass accumulation. The
chosen species – karaka (Corynocarpus laevigatus) (n = 6),
kānuka (Kunzea ericoides) (n = 5), lemonwood (Pittosporum
eugenioides) (n = 6), and kōhūhū (P. tenuifolium) (n = 4) – are
all widely planted in the Auckland Region in revegetation and
restoration projects (Wilcox 2012). Kunzea ericoides can grow
into a canopy tree (up to 20 m) whereas the other species are
smaller trees (up to 15 m) (Wardle 1991).
Due to access problems we were only able to select trees
from certain areas. As far as was possible, individuals typical
of the height and growth form for that species were selected.
Thus, we ignored obviously suppressed individuals and trees
on the edge that often had a different growth form to those
in the interior.
Tree measurements and sample preparation
The trees were measured for their height as well as diameter
at breast height (1.4 m; DBH) above what had been ground
level. The root plate was cut off, and attached soil and stones
carefully removed. In the extraction process, it was inevitable
that fine and medium roots were lost. However, from our
observations, the majority of the major roots were intact; usually
it was the distal ends that were missing. The stem was cut up
into manageable sections and the crown (including branches
and leaves) separated; the crown was defined as the point at
which multi-branching of leaf-bearing branches occurred.
Disks were cut from the stem of each tree for more detailed
laboratory analysis. Each part of the tree was weighed to an
accuracy of 0.1 kg.
Each section of each tree was then chipped using a
commercial tree chipper. Approximate 5-kg samples of
wood chips from the roots, stem and crown of each tree were
retained for laboratory analysis. The approach was to start the
chipping, allow chips to begin to be produced, and then collect
the subsequent 5 kg. This was to avoid any contamination
between samples. The wood chip material was subsampled,
freeze-dried, and then ground to a fine powder using a ultra
centrifugal mill (ZM 200, Retsch GmbH, Haan, Germany).
Organic layer and mineral soil sampling
Organic layer (also called forest floor or O-horizon) and mineral
soil samples were collected beneath 12 trees (n = 3 samples
per tree species). The organic layer was sampled within 20
× 25 cm quadrats and mineral soil samples were taken with
an auger from 0–5 cm depth. Mineral soil samples (only one
sample per tree species) were also taken, at 50–55 cm depth,
to estimate the amount of carbon stored at the time of tree
planting. Soil bulk density was calculated as the quotient
of soil dry weight (105°C, 48 h) and soil volume. The stone
content of the infill material was less than 5%.
The biomass of the organic layer was estimated by
oven-drying the material at 105°C for 48 h after removing
a homogeneous subsample for carbon analysis. Mineral soil
samples and the forest-floor subsamples were oven-dried at
40°C and ground (ZM 200, Retsch GmbH, Haan, Germany).
Mineral soil samples were passed through a 2-mm sieve
before grinding.
Wood density and carbon analysis
The discs, sections of roots, and crown samples were weighed,
air-dried, and reweighed. The volume of the disc and root
sections was determined by water displacement. The samples
were then dried at 105°C until constant weight was achieved
(typically 4 days) to estimate both fresh density and dry wood
density.
Plant and soil material were analysed for carbon
concentration by dry combustion using an elemental analyser
(True Spec, LECO Corporation, St. Joseph, Michigan, USA).
Every 10th sample was analysed twice. The analytical precision
was 1.3% for plant and 0.15% for soil material.
Proc GLM (SAS version 9.2, SAS Institute Inc., Cary,
NC, USA) was used to identify differences in tree and tissue
characteristics among species.
Tree and stand carbon accumulation and sequestration
Tree felling is often not an option to derive tree carbon
stocks, particularly in urban areas. However, in this case
we were provided with mostly complete trees that had been
mechanically extracted from the ground. The amount of carbon
stored in each tree component was estimated by multiplying
carbon concentrations and the mass (dry weight) of a given
component. The total tree carbon is the sum of carbon stored
in above- and below-ground tree components. The proportion
of below-ground carbon to total tree carbon was calculated as
the quotient of root carbon and total carbon stocks.
From these data we were also able to develop an allometric
equation that could be applied to the rest of the trees in
Newmarket Park. We also had other survey data on planted
forests in the Auckland Region, many planted during the same
time period. From these data, species-specific growth curves
were constructed to estimate individual tree heights for the
measured trees.
The amount of carbon stored in the organic layer was
calculated as the product of carbon concentration and the dry
weight of the organic layer. Mineral soil carbon density (Mg C
ha–1) was estimated by multiplying soil carbon concentration
with bulk density and soil depth. The amount of carbon stored
at 50–55 cm depth (carbon density at the time of tree planting
= baseline carbon density) was subtracted from the carbon
density at 0–5 cm depth to calculate the accumulation of soil
carbon since planting.
Schwendenmann, Mitchell: Carbon accumulation in an urban park 215

Results Organic layer and mineral soil carbon

Tree and tissue characteristics and carbon allocation on
a per-tree basis
Comparisons between species are shown in Table 1. Tree
height and dry biomass differed among tree species. Marked
differences were also found in the ratios of fresh to dry weight
and wood densities. Carbon concentration in stem, root,
and crown tissue ranged between 43.3% and 49.6%, with
crown tissue (branches + leaves) tending to have the highest
carbon concentration independent of tree species. The carbon
concentration in all tissue components was highest in Kunzea
ericoides.
Carbon accumulation on a per-tree basis varied
between 15 and 62 kg with Pittosporum eugenioides having
accumulated four times more carbon over the past 27 years than
Corynocarpus laevigatus. The amount of carbon accumulated
in the root biomass accounted for 16–23% of total tree carbon.
This did not relate to any of the other measured variables and
is probably species specific.
Carbon concentration and density (Table 1) in the organic layer
was significantly higher beneath Kunzea ericoides than under
the other tree species. Carbon concentration in the mineral soil
ranged from 5.3% to 7.3%. Carbon density in the top 5 cm
varied between 23 and 32 Mg C ha–1 with the highest amount
of carbon measured under Pittosporum tenuifolium.
Allometric equations based on Newmarket Park data
There are several allometric equations available for estimating
biomass and carbon stocks in natural New Zealand forest
systems (e.g. Coomes et al. 2002; Beets et al. 2012). However,
we initially carried out a standard statistical analysis of the
data without any a priori assumptions as to the nature of the
relationship between tree dimension and tree carbon content.
In particular we were interested in whether we could find an
equation that could be applied to all the species for which we
had biomass data. The best relationship was found using a
polynomial function (r² = 0.8737, F = 38.2, d.f. = 3 and 16,
P < 0.001) of the following form:

Table 1. Summary of comparisons among species for the tree and soil characteristics at Newmarket Park, Auckland, New
Zealand. Values are means ± standard deviation. Significance * P < 0.05; ** P < 0.01; *** P < 0.001; ns – no significant
difference.
__________________________________________________________________________________________________________________________________________________________________

Corynocarpus Kunzea Pittosporum Pittosporum P

laevigatus (n = 6) ericoides (n = 5) eugenioides (n = 6) tenuifolium (n = 4)
__________________________________________________________________________________________________________________________________________________________________

Tree height (m) 6.7 ± 0.5 11.3 ± 0.4 10.5 ± 0.6 9.2 ± 0.4 ***
Tree diameter (cm) 12.0 ± 0.8 12.3 ± 2.3 15.0 ± 0.6 13.6 ± 1.3 *
Tree biomass (kg) 46.2 ± 10.5 89.6 ± 15.3 166.1 ± 25.5 113.6 ± 39.4 **
(dry weight)
Proportion of total biomass
Crown (%) 30.3 ± 2.9 15.8 ± 1.1 20.1 ± 1.4 22.5 ± 2.6 **
Stemwood (%) 51.1 ± 2.6 71.7 ± 1.6 60.1 ± 2.8 62.5 ± 1.9 ***
Roots (%) 18.6 ± 3.0 12.5 ± 1.4 19.7 ± 1.7 15.0 ± 2.3 ns
Root-to-shoot ratio 0.23 ± 0.04 0.19 ± 0.05 0.25 ± 0.09 0.29 ± 0.08 **
Wood density
Stemwood (g cm–3) 0.47 ± 0.01 0.68 ± 0.01 0.63 ± 0.00 0.66 ± 0.01 ***
Roots (g cm–3) 0.50 ± 0.01 0.70 ± 0.01 0.65 ± 0.04 0.71 ± 0.07 **
Carbon concentration
Crown (%) 44.9 ± 1.4 49.6 ± 1.5 46.6 ± 0.4 47.6 ± 0.6 ***
Stemwood (%) 45.2 ± 0.1 47.0 ± 0.3 45.9 ± 0.2 45.6 ± 0.1 ***
Root (%) 43.4 ± 2.4 46.3 ± 1.2 43.8 ± 1.7 44.7 ± 0.5 ns
Tree carbon content (kg tree–1) 15.1 ± 3.2 33.5 ± 5.8 62.1 ± 9.5 42.9 ± 15.0 **
Proportion of total tree carbon
Stemwood + crown (%) 76.7 ± 4.0 84.5 ± 1.7 77.6 ± 1.9 82.2 ± 2.8 **
Roots (%) 23.4 ± 4.0 15.5 ± 1.7 22.4 ± 1.9 17.8 ± 2.8 ***
Organic layer1
Biomass (dry weight) (Mg ha–1) 16.6 ± 3.1 22.5 ± 4.1 15.8 ± 3.5 16.2 ± 4.2 ns
Carbon concentration (%) 37.4 ± 1.6 40.5 ± 0.7 34.2 ± 5.5 33.5 ± 1.5 **
Carbon density (Mg C ha–1) 6.2 ± 0.8 9.1 ± 1.3 5.4 ± 1.6 5.5 ± 1.1 ***
Mineral soil1
0–5 cm
Carbon concentration (%) 5.8 ± 0.4 5.7 ± 0.7 5.3 ± 0.5 7.7 ± 0.4 ***
Carbon density (Mg C ha–1) 25.5 ± 1.7 25.2 ± 3.2 23.3 ± 2.0 32.3 ± 1.5 ***
50–55 cm
Carbon concentration (%) 0.5 0.8 0.8 1.9 ***
Carbon density (Mg C ha–1) 2.9 5.3 5.3 12.9 ***
__________________________________________________________________________________________________________________________________________________________________
1
n = 3 per tree species
216 New Zealand Journal of Ecology, Vol. 38, No. 2, 2014

Tree carbon content (kg tree–1) = −2533.5 volume3 + 1323.2 values for asp based on those presented in Beets et al. (2012).
volume2 + 117.59 volume, where volume (m3) = tree height There is clearly much variability between species for asp, with
× tree basal area. no obvious patterns. However, hardwood and softwood species
tend to have different wood characteristics. In our study we only
This polynomial function was then used to estimate the
had hardwood species thus, for the branches and foliage terms,
carbon storage at the stand level using the transect data.
we used the mean asp based on the hardwood species listed in
Beets et al. (2012). We identified that variation in the asp value
The second allometric equation developed for Newmarket
of the ‘stem and branch carbon’ term has the greatest effect on
Park is based on a power function. This equation has three
overall tree carbon estimation; this also corresponded to our
components (stem and branches, crown and roots) and
‘trunk’ measurement. For each of our species, we developed,
combined these components such that we could estimate
through successive approximation, a ‘stem and branch carbon’
either the above-ground carbon or the total tree carbon. The
asp value, which was used to estimate the carbon content of
equation took the form:
each individual of a species.
Tree carbon content (kg C) = 0.0023DBH3.3885 +
Tree carbon content (kg C) = asp (DBH2 × Height)0.936 +
0.0121DBH2.5276 + 0.009DBH2.4966.
0.0197 DBH0.936 + 0.0148 DBH1.595.
This corresponds to: stem and branches + crown + root.
The asp was modified until the sum of the carbon content
for individuals of that species was equivalent to the measured
Both the polynomial (Proc GLM) and power function (Proc
above-ground value (see Table 2 for the asp values).
POWER) were developed using SAS version 9.2 (SAS Institute
Inc., Cary, NC, USA). The variables were tree volume and
Stand carbon accumulation and sequestration rates
tree diameter, respectively.
The vegetation survey included 12 species that had been
Comparison with the allometric equations of Beets et al. planted. Kunzea ericoides was present in the replanted area but
(2012) not along the four transects. Tree density varied considerably
among species, reflecting differing survival rates or initial
Beets et al. (2012) developed allometric equations (using power
densities (Table 3). Over the 27 years of its existence this forest
functions) for New Zealand forest tree species on the basis
stand accumulated a total 45.9 Mg C ha–1 in above- and below-
of biomass data of approximately 140 trees. The authors also
ground tree biomass (Table 3). The tree carbon sequestration
demonstrate that, for improved precision, parameterisation
rate was 1.7 Mg C ha–1 year–1, with considerable differences
of the species-specific equations varies between species
among species (Table 3). Including the amount of carbon
(Beets et al. 2012). Power functions are the most widely used
stored in the organic layer (6.5 Mg C ha–1) and mineral soil
functions for allometric equations (Picard et al. 2012), but in
(0–5 cm; 20.2 Mg C ha–1 corrected for mineral soil carbon
our analysis we found that a polynomial function provided the
density at the time of tree planting), a total of 72.7 Mg C ha–1
best estimation of total tree carbon content. It is thus important
was stored in this 27-year-old urban park (Table 3).
to compare our results with those that used allometric equations
The ‘Newmarket power function’ and ‘Beets general
based on power functions.
equation’ were also applied to the Newmarket Park vegetation
Presented in Table 2 are results for both above-ground and
survey data to estimate overall carbon accumulation at the
total carbon based on our (destructive) analysis together with
site (Table 4).
a set of additional analyses. The ‘Beets general mixed-species
equation’ (Beets et al. 2012) is expressed as:
Tree carbon content (kg C) = 0.0162 (DBH2 × Height)0.943 Discussion
+ 0.0175 DBH2.2 + 0.0171 DBH1.75.
Carbon concentrations of tree tissues and carbon
The ‘Beets modified equation’ is based on the above equation accumulation on a per-tree basis
but using a species-specific ‘a’ parameter (asp). In our study The higher carbon concentration in crown tissue is explained
we did not separate foliage from branches thus we retained by the high proportion of foliage. Leaves and needles have

Table 2. Comparison of the measured carbon content with the estimated carbon content derived from allometric equations.
a sp = species-specific ‘a’ parameter.
__________________________________________________________________________________________________________________________________________________________________

Measured Polynomial Power Beets general Beets modified asp

__________________________________________________________________________________________________________________________________________________________________

Above-ground carbon (kg C)

Corynocarpus laevigatus 69.8 86.6 111.8 101.3 69.8 0.0161
Kunzea ericoides 143.0 171.9 120.1 139.2 143.1 0.0223
Pittosporum eugenioides 290.9 271.8 260.4 230.4 290.1 0.0283
Pittosporum tenuifolium 144.6 116.5 108.1 104.2 144.2 0.0318
Total (kg C) 648.3 646.7 600.4 575.0 647.1
__________________________________________________________________________________________________________________________________________________________________

Total-tree carbon (kg C) 802.2 799.0 738.9 766.71 805.62

__________________________________________________________________________________________________________________________________________________________________
1
This value is calculated from the assumption that roots represent 25% of the carbon content.
2
This value is calculated from our observation that roots on average comprise 19.8% of tree carbon.
Schwendenmann, Mitchell: Carbon accumulation in an urban park 217

Table 3. Stem density, basal area, carbon accumulation and sequestration of trees, organic layer and mineral soil carbon density (0–5 cm).
Stem density and basal area result from the vegetation survey. The tree carbon values were estimated using the ‘polynomial function’
derived from biomass data of 21 trees felled in Newmarket Park, Auckland, New Zealand.
__________________________________________________________________________________________________________________________________________________________________
1
Species Stem density Basal area Carbon accumulation Carbon sequestration
(ha–1) (m2 ha–1) (Mg C ha–1) (kg C ha–1 year–1)
__________________________________________________________________________________________________________________________________________________________________

Alectryon excelsa 75 1.4 4.19 155.2

Coprosma repens 50 0.6 0.78 28.9
Coprosma robusta 61 0.2 0.07 2.6
Corynocarpus laevigatus 28 0.3 0.22 8.1
Macropiper excelsa 28 0.02 0.002 0.1
Melicytus ramiflorus 301 2.4 3.90 144.4
Myoporum laetum 28 0.1 0.02 0.7
Myrsine australis 25 0.02 0.003 0.1
Pittosporum eugenioides 570 4.8 10.21 378.1
Pittosporum tenuifolium 464 4.0 7.65 283.3
Schefflera digitata 165 1.4 1.48 54.8
Vitex lucens 418 7.1 17.35 642.6
Tree total 2212 22.3 45.9 1699.1
Organic layer 6.5 ± 1.9
Mineral soil
0–5 cm 26.6 ± 4.0
50–55 cm2 6.4 ± 4.0
Mineral soil total 20.2 ± 1.9
Stand total 72.7
__________________________________________________________________________________________________________________________________________________________________
1
Many trees were multi-stemmed, each stem having been measured. This measure was used rather than reporting tree density.
2
Carbon density at the time of planting (= baseline carbon density). This value was subtracted from the carbon density at 0–5 cm depth
to calculate the total stand carbon accumulation since planting.

Table 4. Site estimation of carbon content at Newmarket Park, Auckland, New Zealand. Values derived from three allometric
equations are provided for both individual species and overall estimate of carbon.
__________________________________________________________________________________________________________________________________________________________________

Species Polynomial Newmarket power Beets general

(Mg C ha–1)
__________________________________________________________________________________________________________________________________________________________________

Alectryon excelsa 4.19 3.50 2.54

Coprosma repens 0.78 1.11 0.84
Coprosma robusta 0.07 0.17 0.18
Corynocarpus laevigatus 0.22 0.44 0.33
Macropiper excelsum 0.002 0.01 0.01
Melicytus ramiflorus 3.90 4.54 3.66
Myoporum laetum 0.02 0.05 0.06
Myrsine australis 0.003 0.01 0.02
Pittosporum eugenioides 10.21 9.21 8.43
Pittosporum tenuifolium 7.65 7.77 6.69
Schefflera digitata 1.48 2.70 1.77
Vitex lucens 17.35 17.65 12.04
Total carbon 45.9 47.2 48.81
__________________________________________________________________________________________________________________________________________________________________
1
This allometric equation only estimates above-ground carbon, and has been corrected for an assumed 25% root carbon.

been reported to have higher carbon concentrations compared
with stem wood and roots (e.g. Bert & Danjon 2006; Mello
et al. 2012). Stem-wood carbon concentrations measured in
the four New Zealand native tree species (45.2–47.0%) were
at the lower end of values reported for angiosperms from
across the globe (47.7 ± 0.3%) (Thomas & Martin 2012). In
most New Zealand studies a constant conversion factor of
0.5 is used to convert dry biomass into carbon stocks (e.g.
Beets et al. 2012; Carswell et al. 2012). Using the general
default value of 0.5, the carbon accumulation of individual
trees in Newmarket Park would have been overestimated by
3.0–4.8%. This illustrates the need to obtain species-specific
data on tree tissue carbon.
The amount of carbon accumulated in individual trees
differed considerably among tree species (Table 1). Small
tree size and low wood density explain the low amount of
carbon stored in Corynocarpus laevigatus. Total-tree carbon
accumulation is also influenced by species-specific allocation
strategies (e.g. Campioli et al. 2010). A higher proportion of
stem and root biomass resulted in higher tree carbon stocks in
Kunzea ericoides, Pittosporum eugenioides and P. tenuifolium.
The allocation into slowly decomposable carbon pools such
as stems and roots also implies that the carbon sequestration
capacity of these species in the long term is higher than the
218 New Zealand Journal of Ecology, Vol. 38, No. 2, 2014
carbon sequestration capacity of Corynocarpus laevigatus.
Roots as a percentage of total tree biomass and the root-to-
shoot ratio in our study were in general lower than the values
reported by Phillips and Watson (1994) and Hart et al. (2003).
For example, live fine and coarse roots were 22% of total live
hard beech (Nothofagus truncata, 6–80 cm in diameter) biomass
with a root-to-shoot ratio of 0.28 (Hart et al. 2003). The total
root biomass in six 14-year-old Pinus radiata trees (21–31 cm
in diameter) represented 30% of the total tree biomass and had
a root-to-shoot ratio of 0.43 (Phillips & Watson 1994). Kunzea
ericoides (32 years old, 12.7 cm in diameter) growing in the
East Coast hills had a root biomass (roots > 2 mm) of 20.1
kg tree–1 (Watson et al. 1995) compared with 15.4 kg tree–1at
Newmarket Park. Owing to the lack of species-specific root
biomass data, a root-to-shoot ratio of 0.25, based on the IPPC
guidelines for mature temperate broadleaved forests, is often
used to account for below-ground biomass in New Zealand
trees (e.g. Beets et al. 2012).
Some of the differences in root biomass can be explained
by variations in root extraction methods or data collection
methods. Our lower root biomass and root-to-shoot ratio may
partly be explained by an underestimation of fine roots due to
the root extraction procedure. However, trees grown in urban
environments may allocate less biomass and carbon to roots. A
recent study showed that urban-grown Quercus rubra seedlings
had a lower root-to-shoot ratio than seedlings grown in remote
areas (Searle et al. 2012). The decrease in biomass allocation
to roots has been explained by higher growth temperatures in
urban areas (Searle et al. 2012). Chen et al. (2013) associated
lower fine root biomass in urban Pinus massoniana forests
with higher N depositions in urbanised areas.
Comparison with carbon sequestration rates elsewhere
in New Zealand
Carbon accumulated in above- and below-ground tree biomass
at Newmarket Park (45.9 Mg C ha–1) was within the range
reported for New Zealand shrubland studies. For example,
Coomes et al. (2002) estimated 48.6 ± 13.5 Mg C ha–1 (above-
and below-ground biomass) in Leptospermum scoparium –
Coprosma propinqua dominated shrublands growing north
of Christchurch. An above-ground biomass of 34 Mg C ha–1
was reported for different types of New Zealand shrubland
(Beets et al. 2009 cited in Carswell et al. 2012). Between 12
and 31 Mg C ha–1 were estimated in overseas urban forests
under similar climatic conditions to Newmarket Park (Davies
et al. 2011; Hutyra et al. 2011; Strohbach & Haase 2012).
The sequestration rate of 1.7 Mg C ha–1 year–1 determined
for Newmarket Park was within the range of values reported
for New Zealand shrublands. The rate of carbon sequestration
for young (age < 50 years) Kunzea ericoides and coastal
broadleaved stands was 2.3 Mg C ha–1 year–1 (Carswell et al.
2012). A carbon sequestration rate of 2.1 ± 0.2 Mg C ha–1
year –1 for approximately 40-year-old Leptospermum  ground carbon, underestimating by some 11%. If height data
scoparium – Kunzea ericoides stands across six sites was
reported by Trotter et al. (2005). In contrast, the sequestration
rate of a Pinus radiata plantation forest over a rotation cycle is
about 8 Mg C ha–1year–1 (Maclaren 2000). Our sequestration
rate was lower than the average sequestration rate of 2.8 C
ha–1year–1 by urban trees in the United States (Nowak et al.
2013).
The slightly lower sequestration rates for the planted
forest at Newmarket Park can partly be explained by the
comparatively low tree density. Only 2212 stems ha–1 were
measured in the stand. In contrast, tree density in a naturally
regenerating Kunzea ericoides stand of similar age was 17 330
stems ha–1 (Carswell et al. 2012). Growth rates might also have
been reduced as trees were planted on soils and sediments
characterised by very low carbon and nitrogen contents. Despite
these adverse conditions this planted urban forest sequestered
carbon at a similar rate to natural vegetation in New Zealand.
The amount of carbon stored in the organic layer (5–9
Mg C ha–1) at Newmarket Park was slightly higher than
values measured across 10 urban forests sites (2–6 Mg C
ha–1) in Auckland (Wang 2012). Higher organic (litter)
layer carbon stocks (8.9 ± 2.2 Mg C ha–1) were found in the
Leptospermum scoparium – Coprosma propinqua dominated
shrublands investigated by Coomes et al. (2002). For two
species, Corynocarpus laevigatus and Kunzea ericoides,
the organic layer- and mineral-soil-associated carbon pools
exceeded the amount of carbon stored in tree biomass. Carbon
stored in soil can exceed above-ground carbon and the
importance of soil carbon storage has recently been highlighted
in a study on carbon stock in urban ecosystems (Edmondson
et al. 2012). Our results suggest that the amount of soil organic
carbon is tree species dependent. The effect of tree species on
carbon accumulation in soils is explained by the quality and
amount of litterfall and root residues (e.g. Finzi et al. 1998;
Schulp et al. 2008). The fact that Kunzea ericoides, which had
the highest carbon stock in the organic layer, did not have the
highest tree biomass stocks implies that there are differences
in litter decomposition rates between species.
Comparison of allometric equations
The polynomial function worked well when applied over the
whole Newmarket Park dataset, despite the varying timber
characteristics. However, for individual trees the accuracy
varied considerably. Our polynomial function requires only
diameter and height. As these tree characteristics are easy
to measure, tree carbon storage across a wider scale could
be evaluated. Still, this allometric equation has to be further
improved by measuring additional tree species and sites. Using
this allometric equation for trees outside the diameter range of
trees investigated in this study may not provide reliable carbon
values, as polynomial functions are subject to large errors when
extrapolated outside the measured range (Picard et al. 2012).
The comparison between our measured tree carbon
content and other allometric equations revealed that the more
generalised equations did less well at estimating tree carbon
content than did the more ‘optimised’ equations. The ‘Beets
general mixed-species equation’ with an added 25% for root
carbon provided a good approximation of the measured total-
tree carbon values. This suggests that where both height and
diameter measurements are available, this equation would
be a good approach for non-destructively estimating total
carbon content of a New Zealand mixed forest. Surprisingly
this allometric equation did least well at estimating the above-
are not available, then an allometric equation based on power
functions and diameter may well prove to be reasonably robust
as shown for North American tree species (Jenkins et al. 2004).
The other feature to emerge from these results is that, except
where the equations are species-specific, these equations are
quite imprecise at estimating the carbon content of individual
trees at a site. However, when applied to the overall site
estimation, the applicable equations all provided a similar range
of values (Table 4). It might be expected that the ‘polynomial’
and ‘power’ functions would provide a similar outcome as
they are based on local data. Interestingly, the ‘Beets general
Schwendenmann, Mitchell: Carbon accumulation in an urban park
equation’ produced a similar result, suggesting the wide
applicability of this equation. If many more species can be
parameterised then, clearly, using species-specific allometric
equations will be the optimal solution.
Acknowledgements
The former Auckland City Council provided a small grant
towards this project and facilitated the provision of trees; funds
from The University of Auckland School of Environment PBRF
fund assisted in the completion of this manuscript. We thank
Graham Ussher of Tonkin and Taylor for making it possible to
conduct this project and for organising the on-site contractors;
the unnamed digger drivers who carefully extracted the trees
for us and placed them in a neat stockpile; Peter Crossley for
chainsawing the trees with great skill and efficiency; Gabriella
Ezeta-Ramos for on-site assistance; Sara Howarth of Tonkin
and Taylor for provision of the vegetation description data;
Xiaoshi Zheng for providing her field data on tree heights;
and Bingqin Xu for help with the initial allometric analysis.
We thank Lena Weissert, Fiona Carswell, Chris Lusk, and
an anonymous reviewer for valuable comments on an earlier
version of the manuscript.
References
Auckland Council 2012. The Auckland Plan. http://theplan.
theaucklandplan.govt.nz/ (accessed 15 February 2013).
Beets PN, Kimberley MO, Oliver GR, Pearce SH, Graham
JD, Brandon A 2012. Allometric equations for estimating
carbon stocks in natural forest in New Zealand. Forests
3: 818–839.
Bert D, Danjon F 2006. Carbon concentration variations in
the roots, stem and crown of mature Pinus pinaster (Ait.).
Forest Ecology and Management 222: 279–295.
Campioli M, Gielen B, Granier A,Verstraeten A, Neirynck J,
Janssens IA 2010. Carbon allocation to biomass production
of leaves, fruits and woody organs at seasonal and annual
scale in a deciduous- and evergreen temperate forest.
Biogeosciences 7: 7575–7606.
Carswell FE, Burrows LE, Hall GMJ, Mason NWH, Allen RB
2012. Carbon and plant diversity gain during 200 years of
woody succession in lowland New Zealand. New Zealand
Journal of Ecology 36: 191–202.
Chen H, Zhang W, Gilliam F, Liu L, Huang J, Zhang T, Wang
W, Mo J 2013. Changes in soil carbon sequestration in
Pinus massoniana forests along an urban-to-rural gradient
of southern China. Biogeosciences 10: 6609–6616.
Coomes DA, Allen RB, Scott NA, Goulding C, Beets P 2002.
Designing systems to monitor carbon stocks in forests and
shrublands. Forest Ecology and Management 164: 89–108.
Davies ZG, Edmonson JL, Heinemeyer A, Leake JR, Gaston
KJ 2011. Mapping an urban ecosystem service: quantifying
above-ground carbon storage at a city-wide scale. Journal
of Applied Ecology 48: 1125–1134.
Edmondson JL, Davies ZG, McHugh N, Gaston KJ, Leake
JR 2012. Organic carbon hidden in urban ecosystems.
Scientific Reports 2: 963.
Finzi AC, Van Breemen N, Canham CD 1998. Canopy tree–soil
interactions within temperate forests: species effects on soil
carbon and nitrogen. Ecological Applications 8: 440–446.
Fuller RA, Gaston KJ 2009. The scaling of green space coverage
219
in European cities. Biology Letters 5: 352–355.
Gaffin SR, Rosenzweig C, Kong AYY 2012. [Correspondence]
Adapting to climate change through urban green
infrastructure. Nature Climate Change 2: 704.
Hart PBS, Clinton PW, Allen RB, Nordmeyer AH, Evans G
2003. Biomass and macronutrients (above- and below-
ground) in a New Zealand beech (Nothofagus) forest
ecosystem: implications for carbon storage and sustainable
forest management. Forest Ecology and Management
174: 281–294.
Hutyra LR, Yoon B, Alberti M 2011. Terrestrial carbon stocks
across a gradient of urbanization: a study of the Seattle,
WA region. Global Change Biology 17: 783–797.
Jenkins JC, Chojnacky DC, Heath LS, Birdsey RA 2004.
Comprehensive database of diameter-based biomass
regressions for North American tree species. General
Technical Report NE-319. Newton Square, PA, USDA,
Forest Service, Northeastern Research Station. 45 p.
Maclaren JP 2000. Plantation forestry: its role as a carbon
sink, In: Bouma WJ, Pearman GI, Manning MR eds
Greenhouse: coping with climate change. Collingwood,
Victoria, Australia, CSIRO. Pp. 417–436.
Mello AA, Nutto L, Weber KS, Sanquetta CE, Monteiro de
Matos JL, Becker G 2012. Individual biomass and carbon
equations for Mimosa scabrella Benth. (bracatinga) in
southern Brazil. Silva Fennica 46: 333–343.
NIWA 2012. Annual Climate Summary 2012. NIWA
National Climate Centre. National Institute of Water and
Atmospheric Research. http://www.niwa.co.nz/climate/
summaries/annual/annual-climate-summary-2012
(accessed 15 February 2013).
Nowak DJ, Greenfield EJ, Hoehn RE, Lapoint E 2013. Carbon
storage and sequestration by trees in urban and community
areas of the United States. Environmental Pollution 178:
229–236.
Oldfield EE, Warren RJ, Felson AJ, Bradford MA 2013.
Challenges and future directions in urban afforestation.
Journal of Applied Ecology 50: 1169–1177.
Phillips CJ, Watson AJ 1994. Structural tree root research in
New Zealand: a review. Landcare Research Science Series
7. Lincoln, Manaaki Whenua Press. 71 p.
Picard N, Saint-André L, Henry M 2012. Manual for building
tree volume and biomass allometric equations: from field
measurement to prediction. Rome, Food and Agricultural
Organisation of the United Nations, and Montpellier,
Centre de Coopération Internationale en Recherche
Agronomique pour le Développement (CIRAD). 215 p.
Pouyat RV, McDonnell MJ, Pickett STA 1995. Soil
characteristics of oak stands along an urban-rural land-use
gradient. Journal of Environmental Quality 24: 516–526.
Schulp CJE, Nabuurs GJ, Verburg PH, de Waal RW 2008.
Effect of tree species on carbon stocks in forest floor and
mineral soil and implications for soil carbon inventories.
Forest Ecology and Management 256: 482–490.
Searle SY, Turnbull MH, Boelman NT, Schuster WSF, Yakir
D, Griffin KL 2012. Urban environment of New York
City promotes growth in northern red oak seedlings. Tree
Physiology 32: 389–400.
Strohbach MW, Haase D 2012. Above-ground carbon storage
by urban trees in Leipzig, Germany: Analysis of patterns
in a European city. Landscape and Urban Planning 104:
95–104.
Thomas SC, Martin AR 2012. Carbon content of tree tissues:
a synthesis. Forests 3: 332–352.
220 New Zealand Journal of Ecology, Vol. 38, No. 2, 2014
Trotter C, Tate K, Scott N, Townsend J, Wilde H, Lambie S,
Marden M, Pinkney T 2005. Afforestation/reforestation
of New Zealand marginal pasture lands by indigenous
shrublands: the potential for Kyoto forest sinks. Annals
of Forest Science 62: 865–871.
US EPA 2011. Evaluation of urban soils. Suitability for green
infrastructure or urban agriculture. EPA Publication No.
905R11003. United States Environmental Protection
Agency. 22 p.
Editorial Board member: Chris Lusk
Received 3 May 2013; accepted 16 November 2013
Wang N 2012. Auckland urban forest soil carbon density
and the factors causing its variability. Unpublished MSc
thesis, The University of Auckland, New Zealand. 79 p.
Wardle P 1991. Vegetation of New Zealand. Cambridge, UK,
Cambridge University Press. 672 p.
Watson A, Marden M, Rowan D 1995. Tree species performance
and slope stability. In: Barker DH ed Vegetation and slopes:
stabilisation, protection and ecology. London, Thomas
Telford. Pp. 161–171.
Wilcox MD 2012. Auckland’s remarkable urban forest.
Auckland. Auckland Botanical Society Inc.
Copyright of New Zealand Journal of Ecology is the property of New Zealand Ecological
Society, Inc. and its content may not be copied or emailed to multiple sites or posted to a
listserv without the copyright holder's express written permission. However, users may print,
download, or email articles for individual use.